Advances in Statistical Methods for Natural Language Processing 2011

Instructor

Assistant professor, Mamoru Komachi <komachi--at--is.naist.jp>

Course description

This course gives you a brief introduction to statistical natural language processing in comparison to previous heuristic methods.

You will see several recent statistical approaches to natural language processing, including:

• Statistical machine translation
• Statistical Japanese text input method
• Information retrieval and information extraction.

Dates

1. 9:20-10:50, Dec 16, 2011: Introduction to statistical natural language processing (slides)
2. 9:20-10:50, Dec 20, 2011: Statistical machine translation (slides)
3. 9:20-10:50, Jan 6, 2012: Statistical Japanese text input method (slides)
   • (kiyoshi-k) Hal Daume III and Daniel Marcu. A Noisy-Channel Model for Document Compression. ACL 2002. (slides)
4. 9:20-10:50, Jan 10, 2012: Information retrieval and information extraction
   • (komachi) Kei Uchiumi, Mamoru Komachi, Keigo Machinaga, Toshiyuki Maezawa, Toshinori Satou and Yoshinori Kobayashi. Japanese Abbreviation Expansion with Query and Clickthrough Logs. IJCNLP 2011. (slides)
   • (haruka-m) Pallavi Choudhury, Chris Quirk, and Hisami Suzuki. From pecher to pêcher ... or pécher: Simplifying French Input by Accent Prediction. WTIM 2011. (slides)
   • (kou-j) Zhen Chen and Kai-Fu Lee. A New Statistical Approach to Chinese Pinyin Input. ACL 2000. (slides)

All slides can be accessible from naist.jp.

Venue

L2 (Lecture room 2), Graduate School of Information Science

Grading

100% on your final presentation (either on Jan 6 or 10). You need to choose one research paper relevant to this course, and will explain the paper using slides (you can take up to 15 minutes). Your paper must be somewhat related to noisy channel model.

Paper list (but not limited to)

1. Hal Daume III and Daniel Marcu. A Noisy-Channel Model for Document Compression. ACL 2002.
2. Kevin Knight and Jonathan Graehl. Machine Transliteration. ACL 1997.
3. Zheng Chen and Kai-Fu Lee. A New Statistical Approach to Chinese Pinyin Input. ACL 2000.
4. Yabin Zheng, Chen Li, and Maosong Sun. CHIME: An Efficient Error-Tolerant Chinese Pinyin Input Method. IJCAI 2011.
5. Pallavi Choudhury, Chris Quirk, and Hisami Suzuki. From pecher to pêcher ... or pécher: Simplifying French Input by Accent Prediction. WTIM 2011.
6. Eric Brill and Robert C. Moore. An Improved Error Model for Noisy Channel Spelling Correction. ACL 2000.
7. Mark D. Kernighan, Kenneth W. Church, and William A. Gale. A Spelling Correction Program Based on a Noisy Channel Model. COLING 1990.
8. Kristina Toutanova and Robert Moore. Pronunciation Modeling for Improved Spelling Correction. ACL 2002.
9. Mark Johnson and Eugene Charniak. A TAG-based noisy channel model of speech repairs. ACL 2004.
10. Randy West, Y. Albert Park, and Roger Levy. Billingual Random Walk Models for Automated Grammar Correction of ESL Author-Produced Text. EduNLP 2011.
11. John Lee and Stephanie Seneff. Automatic Grammar Correction for Second-Language Learners. InterSpeech 2006.

All students who want to get a credit for this course must submit their preference of the papers (up to three) to komachi@is by 19:59, Dec 22, 2011.

Office hours

• 15:00-17:00, Dec 22, 2011
• 15:00-17:00, Jan 5, 2012

No appointment needed. Please come by A705. Other hours available upon request.

References

• Philipp Koehn. 2009. Statistical Machine Translation. CUP.
• Daniel Jurafsky and James H. Martin. 2008. Speech and Language Processing. Pearson Prentice Hall.